ARIZONA STATE UNIVERSITY SCHOOL OF SUSTAINABLE ENGINEERING AND THE BUILT ENVIRONMENT Comparative Life Cycle Assessment of Lamps Used in a Classroom at Arizona State University Parker Helble, Elena Hoff, Andy Stein, and Jessica Welch ASU-SSEBE-CESEM-2014-CPR-003 Course Project Report Series June 2014     Comparative  Life  Cycle  Assessment  of   T8  Lamps  Used  in  a  Classroom  at   Arizona  State  University   Life  Cycle  Assessment  for  Civil  Systems  –  Dr.  Chester   SOS  598  :  Spring  2014         Final  Report         Completed  by:   Parker  Helble,  Elena  Hoff,  Andy  Stein,  and  Jessica  Welch     2   Table of Contents Background: ............................................................................................... 2 Introduction: ............................................................................................... 3 Methodology: .............................................................................................. 3 Results: ....................................................................................................... 9 Uncertainty: .............................................................................................. 12 Conclusion:............................................................................................... 14 References: ............................................................................................... 15 Appendix: .................................................................................................. 16 Background: A typical classroom at Arizona State University uses T8 fluorescent light bulbs. Newer technologies include an LED T8 equivalent. Fluorescent bulbs use mercury vapor to produce visible light, and LED bulbs contain light-emitting diodes. A plethora of life cycle assessments have been conducted to compare different types of light bulbs. However, no existing LCAs which specifically analyze fluorescent and LED T8 lamps could be found. The goal is to fill this gap by analyzing the environmental and economical impacts of these different lamps in a specific scenario relevant to Arizona State University. 3 Introduction: The ultimate goal of this LCA is to give Arizona State University specific advice on possible changes in lighting systems that will reduce environmental impacts and support ASU’s sustainability efforts. The aim is to assess the potential for a decrease in specific environmental impacts (CO2 emissions and energy use) and economic impact (cost) from changing to a different type of lighting in a prototypical classroom in Wrigley Hall. The scope of this assessment is to analyze the impacts of T8 lamps lasting 50,000 hours. Thus, a functional unit was defined as 50,000 hours of use, maintaining roughly 825 lumens. To put this in perspective, 50,000 hours is equivalent to 8 hours of use per day, 365 days per year, for approximately 17.1 years. An attributional LCA was conducted comparing two T8 lamps: a fluorescent 32 watt bulb, and an LED 22 watt bulb. Due to a lack of existing data on the T8 lamps, data from existing LCAs of screw-base CFL 15 watt lamps and LED 12.5 watt lamps was used for the raw material extraction and manufacturing phases. Actual LCI data on the T8s was used for calculation in the use phase of the Life Cycle Assessment. The indicators chosen were Global Warming Potential (GWP), measured in kg CO2 eq., energy, measured in kWh, and cost, measured in US dollars. Methodology: The methodology used in this LCA included the following steps, which are outlined below. Each step is elaborated further within this section: Step 1: Defined the scope and system boundary of the study. Step 2: Selected two types of T8 lamps for comparison in this study and defined a functional unit. Step 3: Defined three indicators and used these to compare the two lamps. Step 4: Collected Life Cycle Inventory (LCI) data from existing LCA studies of screwbase compact fluorescent and LED lamps for cradle-to-gate phases. Step 5: Converted the LCI data for screw-base lamps to data for T8 lamps by multiplying by a mass ratio factor. Step 6: Calculated the use phase of each T8 lighting option based on its energy consumption and associated impacts. Step 7: Combined the LCI data gathered and calculated from Steps 4 and 5 to provide relevant data for analysis. Step 1: Scope and System Boundary The goal of this LCA study is to understand all that goes into the lighting of one basement room in Wrigley Hall each day. A basement classroom was selected in order to hold the amount of natural light constant at 0 lumens. The focus of this study is narrowed to scope of the impacts of just one light bulb for a certain period of time. This includes impacts associated with the lamps only, and not those associated with the fixtures, electrical hook ups, or other necessary lighting components such as ballasts. Not only does this study consider the production of each lamp, but also the electricity used to power the lamp, how the power is generated and the impacts of that generation. After the impacts from each individual lamp are assessed and analyzed, the results are compiled in order to discuss the impacts of an entire classroom. 4 The system boundary (illustrated in Figure 1 and described in Table 1) for this LCA includes three life cycle phases: raw material extraction, manufacturing and use. It is a cradleto-use analysis while disregarding the transportation phase. Figure 1: System Boundary used throughout the LCA Table 1: Summary of Life Cycle Phases Used in the LCA Life Cycle Phase: Inputs (Raw materials extraction) Manufacturing (Processing and assembly of lamps) Included? Summary of Phase ✓ This phase accounts for the emissions and resource usage resulting from the extraction of the raw materials needed to construct each lamp. Raw materials extraction associated with power generation for the electricity is not considered within the system boundary. ✓ The manufacturing phase considers the processes involved in manufacturing the lamp out of the raw materials and the assembly of the lamp at the manufacturing site. It considers the energy used throughout the process of manufacturing the lamp, as well as the emissions associated with fabricating each lamp. 5 Transportation (Transportation from manufacturing site to end user) Use (Use of lamp by end user) End of life (Disposal/recycling of lamps after use) X ✓ X The phase of the life cycle where each lamp is distributed from the manufacturer to the end user is not considered. Compared to other life cycle stages, the impacts associated with the transportation and distribution stages are negligible1. The energy use, CO2 emissions and cost associated with this phase are calculated based upon the electricity consumed by 50,000 hours of use. This phase of the life cycle is not included in the system boundary. Lamps containing mercury are considered hazardous waste material and so the end of life choices for disposal or recycling are left up to the discretion of each end user and their regional regulations2. Under consideration for the scope of this LCA, it was determined that this phase in the life cycle of each light bulb could be determined by ASU’s policies and did not need to be included for this study. Step 2: Lamp Selection and Definition of Function Unit This project is a comparative study, therefore two different lighting technologies were chosen for analysis. Fluorescent T8 lamps rated at 32 watts, which are currently used in the sample classroom, are being compared in this study to the possibility of 22 watt LED T8 replacements. Since published LCI data on T8 lamps is scarce this study chose to find data on a screw-base 15 watt CFL and a 12.5 watt LED lamp for the raw material extraction and manufacturing phases and then extrapolated that data for use in this study. The average lifetime of one 22 watt LED T8 lamp - 50,000 hours3 - was used as the functional unit. The average lifetime of one 32 watt fluorescent T8 lamp is 30,000 hours4. Therefore, two fluorescent T8 lamps must be manufactured and put into use in order to reach the lifetime of one LED T8 lamp. Step 3: Defined Indicators The three indicators for measuring environmental or economical impacts of each lighting option in this study are Energy, measured in kWh, Global Warming Potential (GWP), measured in kg CO2 eq., and Cost, measured in US Dollars. It is believed that these three impacts are the most relevant to ASU and will allow the responsible parties to make the best decision regarding which type of lamp to use in the future. 1 US Department of Energy, 2012b, p. 13 US Environmental Protection Agency, n.d. Premier Lighting, 2013 4 Premier Lighting, 2013 2 2 3 6 Step 4: Collected Life Cycle Inventory (LCI) The following information is used to gather data for the cradle-to-gate part of the study for the first two indicators, Energy and GWP. Due to a lack of data on T8 lamps, existing Life Cycle Inventory data was used for screw-base lamps instead. Actual data on T8 lamps was used for the cost indicator. In order to maintain a consistent period of operation two fluorescent T8 lamps need to be produced to match the lifetime of one LED T8 (see step 2). Therefore, the final data for the screw-base CFL lamps and fluorescent T8 lamps is multiplied by 2. It is assumed that the screw-base lamps have the same lifetimes as their T8 counterparts (30,000 hours for fluorescent, and 50,000 hours for LED). Indicator 1: Energy CFL: Using a conversion factor of 6.9x10-4 metric tons of CO2 / kWh5 for raw material extraction, the global warming potential associated with the raw material extraction for the production of screw-base compact fluorescent lamps (10.68 kg CO2 eq6) was converted to the equivalent energy usage of 15.478 kWh. During the manufacturing life-cycle phase 170 MJ/20 million lumen hours7 was converted to 15.58 kWh (See Table 11 in Appendix). Thus, for one CFL screw-base lamp, the total energy used during the raw material extraction and manufacturing phases is 31.058 kWh x 2 (lamps produced) = 62.116 kWh. LED: Using the same conversion factor, the global warming potential associated with the raw material extraction for the production of screw-based LED lamps (12.752 kg CO2 eq8) as converted to the equivalent energy usage of 18.481 kWh. During the manufacturing lifecycle phase 343 MJ/20 million9 lumen hours was converted to 95.28 kWh (See Table 11 in Appendix). Thus, for one LED screw-base lamp, the total energy used during raw material extraction and manufacturing is 113.761 kWh. Indicator 2: GWP CFL: The global warming potential associated with raw material extraction of a compact fluorescent lamp is 10.68 kg CO2 eq10 x 2 (lamps produced) = 21.36 kg CO2 eq. The global warming potential associated with manufacturing of a compact fluorescent lamps is 16.56 kg CO2 eq11 x 2 (lamps produced) = 33.12 kg CO2 eq. Thus, for one CFL screw-base lamp, the total GWP for raw material extraction and manufacturing is 54.48 kg CO2 eq. LED: The global warming potential associated with raw material extraction of a LED screw-base lamp is 12.752 kg CO2 eq12. The global warming potential associated with manufacturing a 5 US Department of Energy, 2012a, p. 43 US Department of Energy, 2012b, p. 50, Table 7-2 7 US Department of Energy, 2012a, p. 30, Table 4.5 8 US Department of Energy, 2012b, p. 50, Table 7-3 9 US Department of Energy, 2012b, p. 50, Table 7-2 10 US Department of Energy, 2012b, p. 50, Table 7-2 11 US Department of Energy, 2012b, p. 50, Table 7-2 12 US Department of Energy, 2012b, p. 50, Table 7-3 6 7 LED screw-base lamp is 3.45 kg CO2 eq13. Thus, for one LED screw-base lamp, the total GWP associated with raw material extraction and manufacturing is 16.202 kg CO2 eq. Indicator 3: Cost It is assumed that the purchasing price of each lamp from a retailer includes the cost of raw material extraction and manufacturing. Average prices for both compact fluorescent and LED T8 lamps were used, $5 and $6514, respectively. However, two fluorescent T8 lamps need to be purchased in order to reach 50,000 hours of use, so the price of the fluorescent T8 lamps is actually $10. Step 5: LCI Data Conversion The average mass of both a screw-base CFL and an LED lamp was calculated by using the information for CFL-1 and LED-1 in the EPA’s LCA15. Based on this information, the average mass of a screw-base CFL bulb is 58.1 grams, and the average mass of a screw-base LED bulb is 179 grams. The mass of a fluorescent T8 lamp is 225 grams, and the mass of an LED T8 lamp is 680 grams16. For the raw materials extraction and manufacturing phases, it was assumed that impacts are proportional to mass. Thus, the final data for the screw-base lamps (see below) was multiplied by 3.873 for the CFL lamps and 3.799 for the LED lamps in order to calculate impacts for fluorescent T8 lamps and LED T8 lamps, respectively. These conversions by mass are displayed in Table 2. The cost indicator is not affected by this mass ratio. Table 2: Energy and GWP for the Raw Materials Extraction and Manufacturing Phases of the Life Cycle Energy for Screw-base lamp Mass Ratio Energy for T8 lamp Fluorescent T8 62.116 kWh 3.873 240.575 kWh LED T8 113.761 kWh 3.799 432.178 kWh GWP for Screw-base lamp GWP for T8 lamp Fluorescent T8 54.48 kg CO2 eq. 3.873 211.001 kg CO2 eq LED T8 16.202 kg CO2 eq 3.799 61.551 kg CO2 eq Step 6: Use Phase Calculations Indicator 1: Energy The energy consumed by each lamp was calculated during the use phase based on the watts each lamp is rated for (32 watts for fluorescent T8s, 22 watts for LED T8s) and throughout US Department of Energy, 2012b, p. 50, Table 7-3 Premier Lighting, 2013 15 US Department of Energy, 2013, p. 35, Table 5-9 16 Osram Sylvania Warehouse Personnel, personal communication, February 27, 2014 13 14 8 the functional unit of 50,000 hours. Fluorescent T8s use 1600 kWh and LED T8s use 1100 kWh. Indicator 2: GWP The energy calculations from Indicator 1 were used along with the Arizona New Mexico (AZNM) subregion CO2 output emission rate from the EPA, 1177.61 lbs CO2 eq / MWh17, to determine the use phase CO2 emissions for both types of lamps. Thus it was calculated that the global warming potential during the use phase is 855.42 kg CO2 eq for fluorescent T8 lamps and 588.10 kg CO2 eq for LED T8 lamps. Indicator 3: Cost In order to determine the price per kWh, the average monthly price of electricity was obtained using the SRP Standard Price Plan18, which is used by commercial, business, and professional customers. This average was calculated to be $0.0924 per kWh. The cost of the use phase for each lighting option was then calculated by multiplying the energy used by this price. The operational cost of a fluorescent T8 was found to be $147.84 where as the operation cost of a LED T8 was $101.64. Step 7: Aggregation of Data The gathered LCI data was combined from Step 4 and Step 5, and was added to the use phase calculations in Step 6 for all three indicators to obtain the total life cycle impacts of each type of lamp (Tables 3-5). Table 3: Energy Use (kWh) per 50,000 hours Raw Materials Extraction Manufacturing Use Total Fluorescent T8 119.892 120.682 1600 1840.575 LED T8 70.209 361.969 1100 1532.178 Table 4: Global Warming Potential (kg CO2 eq.) per 50,000 hours Raw Materials Extraction Manufacturing Use Total Fluorescent T8 82.727 128.274 855.416 1066.417 LED T8 48.445 13.107 588.098 649.650 Table 5: Cost (U.S. Dollars) per 50,000 hours Fluorescent T8 Price of bulb(s) (represents raw material extraction and manufacturing) Use Total 17 18 LED T8 10 65 147.84 157.84 101.64 166.64 US Environmental Protection Agency, 2014, p. 1 Salt River Project Agricultural Improvement and Power District, 2012, p. 3 9 Results: As previously mentioned, the three indicators of interest are energy use, global warming potential, and economic cost. These indicators are based on cradle-to-use data, with a functional unit of 50,000 hours of roughly 825 lumens per lamp assuming quality of light is the same. For each indicator, the data provided is calculated for one T8 LED lamp (equivalent to one LED lifetime of 50,000 hours) and two fluorescent T8 lamps because two of them will be required to last for a lifetime of at least 50,000 hours. Indicator 1: Energy Use The largest impacts on energy come from the use of the lamps. Fluorescent T8 lamps consume more energy than LED lamps per 50,000 hours of use. The extraction of raw materials for fluorescent lamps consumes more energy than for LED lamps, because two lamps need to be produced in order to last for 50,000 hours. However, the manufacturing of one LED lamp consumes more energy than manufacturing two fluorescent lamps. Overall, fluorescent lamps consume more energy than LED lamps from raw material extraction through use. Table 6 and Figure 2 show the results for this indicator. Table 6: Total energy use for each lighting option. Figure 2: Total energy consumption for each lighting option, per 50,000 hours of use 10 Indicator 2: Global Warming Potential Similar to energy, the largest impacts on global warming potential come from the use phase. This is largely due to the energy mix for this region in Arizona. Fluorescent T8 lamps produce more kg CO2 eq. than LED lamps do per 50,000 hours of use. Raw material extraction and manufacturing both produce more kg CO2 eq. for fluorescent lamps than for LED lamps as well. Overall, fluorescent lamps produce more kg CO2 eq. than LED lamps do from raw material extraction through use. Table 7 and Figure 3 show the results for this indicator. Table 7: Total global warming potential for each lighting option. Figure 3: Global warming potential for each lighting option 11 Indicator 3: Cost The current retail price of a light bulb is the largest impact for LED lamps. Fluorescent lamps are much less expensive than LED lamps. Even though two fluorescent lamps are needed to reach 50,000 hours, the two fluorescent lamps are still less expensive than one LED lamp. However, the cost of using the lamps for 50,000 hours is higher for fluorescents than for LEDs. Overall, LED lamps cost more than fluorescent lamps. However, these results are extremely sensitive to the price of the lamps. For instance, if the LED lamps only cost $40 instead of $65, the results would flip, and LED lamps would cost less overall than fluorescent lamps. Table 8 and Figure 4 show the results for this indicator. Table 8: Total cost for each lighting option . Figure 4: Total cost for each lighting option, per 50,000 hours use 12 Uncertainty: There are two methods to address uncertainty within this project: uncertainty within the methodology and uncertainty within the data collected. 1. Uncertainty within the methodology: There are many assumptions within the report that create an abundance of uncertainty. For instance, the lack of available life cycle impact assessment data on T8 fluorescent and LED lamps required us to approximate these findings. The procedure to do so is based on acquiring data for screw-base lamps and then extrapolating that data for use in our study based on mass proportionality between screw-base lamps and their T8 equivalents. Therefore, we are assuming that the impacts associated with one kilogram of the screw-base CFL are the same for one kilogram of the T8 fluorescent lamp. This assumption that the impacts we are researching are proportional by mass from the screw-base lamps to their T8 equivalents is a large uncertainty. It is very probable that T8 equivalents require a completely different manufacturing process, thus resulting in differing impacts associated with that phase. Also, it is known that the LED T8s have separate components to their fixture, and require an attached ballast system to function properly. The mass of this ballast was absorbed into the mass of the individual lamp to simplify the mass conversion calculations. This assumption was used to simplify the results for the purpose of this study, so impacts associated with raw material extraction and manufacturing of the ballast component were ignored. As previously mentioned, the end of life phase for each lamp type was not taken into account. If end of life data were to be included in this study, the overall impacts for the three indicators would likely increase. The intentional exclusion of the end of life impacts is a point of uncertainty that has the potential to change the results to favor one lighting option more strongly than the other. 2. Uncertainties within the data collected: All of the data used in the study is coming from secondary data sources and it is assumed that the data used is as accurate as possible for the purpose of this study. The data sources used for this study include some uncertainty because the data was not collected first hand by the team. Because the uncertainty of the data cannot be quantified, a pedigree matrix was used to assess uncertainty. The template used for scoring each uncertainty aspect in the pedigree matrix19 is included in Table 10 in the appendix. The scores assigned for each aspect of uncertainty are shown in Table 9 below. 19 Chester, M., 2014, p. 24 13 Table 9: Data Quality Assessment Pedigree Matrix As explained above, the main source of uncertainty in our project came from the mass equivalency assumption that is explained in the methodology. This is reflected in the high scores that we assigned to the “technological correlation” and “range of variation” indicators. The remaining indicators were all scored as either 1 or 2, indicating that there is not much room for improvement in those categories. If reductions in uncertainty are to be made, the best place to start would be in the mass equivalency assumption. If there were existing LCAs on Fluorescent and LED T8s this data would greatly reduce the uncertainty in the study. 14 Conclusion: According to Sustainability Operations at ASU, “ASU takes seriously reducing consumption, maximizing efficiency and rethinking products and actions20.” The results from this study prove to be very useful to ASU when looking at ways to maximize efficiency and move toward more sustainable operations. Results show that LED T8 lamps use less energy and have less of an impact on global warming potential for 50,000 hours of use compared to fluorescent T8 lamps. In order to translate this data into useable information for ASU decision making, this study is applied to a prototypical ASU classroom. Specifically, this study used Wrigley Hall L114, where there are currently 36 lamps used to sufficiently light the room. Therefore, all of the indicators that are measured per lamp must be multiplied by a factor of 36 to get the total impact from illuminating one classroom. Switching to LED T8 lamps can lead to widespread effects across the campus, and may even be able to help in changing lamp use and purchasing policy. These effects determined within the scope of this study may include a 40% decrease in GWP and a 17% decrease in energy consumption. However, cost is estimated to increase by 6%. Although LED T8s are more expensive than fluorescent T8s, it is still recommended that ASU transition to LED T8s. Over the lifetime of the LED T8 lamps, the savings in energy use will help offset the higher cost of the new lamps. In addition, if ASU is able to receive bulk purchasing rates for the LED T8s, the cost indicator may switch to show that LEDs are more cost efficient. The cost indicator is very sensitive as previously mentioned, and this may be a deciding factor in ASU’s decision. According to Nick Brown, Director of University Sustainability Practices, ASU does not have an overarching body that regulates lamps purchased for campus facilities21. It is recommended that ASU creates a standard purchasing policy that will allow for uniformity in implementation of LED T8s in campus facilities. Going forward, this study can be used as a starting point for analyzing the entire cradle to grave life cycle for both LED and fluorescent T8s. 20 21 Arizona State University Global Institute of Sustainability, n.d. Personal communication with Nick Brown, February 28, 2014 15 References: Arizona State University Global Institute of Sustainability. (n.d.). Sustainability in Practice. Retrieved from http://sustainability.asu.edu/practice/ Chester, M. (2014). Uncertainty. Lecture from SOS 598: Life Cycle Assessment for Civil Systems, April 3, 2014. Premier Lighting. (2013). Should You Replace Your T8 Fluorescent Lamps with T8 LED Tubes? Retrieved from http://www.premierltg.com/should-you-replace-your-t8fluorescent-lamps-with-t8-led-tubes/ Salt River Project Agricultural Improvement and Power District. (2012). Standard Price Plan for General Service. Retrieved from http://www.srpnet.com/prices/pdfx/BusE36.pdf US Department of Energy. (2012a). Life-Cycle Assessment of Energy and Environmental Impacts of LED Lighting Products, Part 1: Review of the LifeCycle Energy Consumption of Incandescent, Compact Fluorescent, and LED Lamps. Retrieved from http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/2012_LED_Lifecycle _Report.pdf US Department of Energy. (2012b). Life-Cycle Assessment of Energy and Environmental Impacts of LED Lighting Products, Part 2: LED Manufacturing and Performance. Retrieved from http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/2012_led_lca-pt2.pdf US Department of Energy. (2013). Life-Cycle Assessment of Energy and Environmental Impacts of LED Lighting Products, Part 3: LED Environmental Testing. Retrieved from http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/2013_led_lcapt3.pdf US Environmental Protection Agency. (n.d.). Frequent Questions about Regulations that Affect the Management and Disposal of Mercury-Containing Light Bulbs (Lamps). Retrieved from http://www.epa.gov/osw/hazard/wastetypes/universal/lamps/faqs.htm US Environmental Protection Agency. (2014). eGRID 9th edition Version 1.0 Year 2010 Summary Tables. Retrieved from http://www.epa.gov/cleanenergy/documents/egridzips/eGRID_9th_edition_V10_year_2010_Summary_Tables.pdf 16 Appendix: Table 10: Data quality assessment pedigree matrix scoring guidelines Table 11: Calculation Spreadsheet 17